Results challenge conventional wisdom about where the brain begins processing visual information

Now, however, a brain imaging study -- published online by the journal Nature Neuroscience on Mar. 2 -- challenges this basic assumption. Using high-resolution functional magnetic resonance imaging (fMRI), a team of researchers from Vanderbilt and Boston universities, have discovered that more complex processing occurs in the initial stages of the visual system than previously thought. Specifically, they have found evidence of processing in the human lateral geniculate nucleus (LGN), a small node in the thalamus in the middle of the brain that relays nerve impulses from the retina to the primary visual cortex.

An important function of the visual cortex is the processing of rudiments of shape, the angles of lines and edges, which are important for defining the outlines of objects. The researchers found that the human LGN is also sensitive to the orientation of lines and that this effect is enhanced when a person simply pays attention to the orientations in an image.

"Our results demonstrate that even the simplest brain structures may play a fundamental role in complex neural processes of perception and attention," said Frank Tong, professor of psychology at Vanderbilt, who conducted the study with postdoctoral fellow Michael Pratte and Sam Ling at Boston University. "They also highlight how higher cortical areas can influence and modulate how we see by modifying the responses of neurons at the earliest stages in the visual pathway through feedback connections."

"The findings challenge the conventional wisdom about how and where in the brain the processing of visual orientation information first occurs," commented Michael A. Steinmetz, acting director of the Division of Extramural Research at the National Eye Institute, which provided funding for the study. "The research also underscores the concept that the perception of visual stimuli evolves from dynamic processes in widely distributed networks in the brain."

The research was supported by National Institutes of Health grants R01 EY01782 and R01 EB000461 and NIH Fellowship F32-EY022569.